Musings on the AI apocalypse

I was watching the video video below and felt like waxing philosophic for a moment. 

I usually hate it when people ask me about the the robot uprising. Is a narrow and misconstrued question that doesn't take into account that we know very little about artificial intelligence or intelligence in general. And we don't know what we don't know. 

The fear from experts is brought on by the fact that the technologies are accelerating. The fact that processing will be twice what it is currently in a just a couple of years, and so on, creates an unknowable horizon of what will be possible. That all being said, we are currently no where near to a dangerous AI. The "awesome" systems of "AI" and "Machine Learning," propped up by bloggers and used by Google, are based on algorithms from the 60's that can be fooled into thinking a turtle is a gun (MIT Paper from last year). 

But that doesn't mean that they are not dangerous. We are in a Jurassic Park situation where people will get so excited "to see if they could, they won't stop to see if they should." And that can lead to an AI apocalypse detailed in the video below. A situation where a single-function stupid-supersmart program finds a "unique" solution to a problem. It is not a malicious Terminator, it is simply machine error.

Musings on STEM: Coding is a Tool

Very recently I had a colleague mention that students should learn coding much as one would learn a second language. At first I liked the analogy. A second language is something that can be valuable, but is ultimately optional. Which is largely how I view coding. But something about it seemed forced and seemed to give coding a higher platform than I like to give it most days. After a little soul-searching I came up with with the premise of this post. Coding is a tool that should be taught as a trade.

Computer code is pervasive. The average american citizen comes into a contact with a computer, running code, dozens of times a day. (And for those with Facebook, hundreds of times). This has helped perpetuate the illusion, yes Illusion, that code has a higher value than other components of life. But exposure does not translate to value. After all, most only eat 3 times a day, and drink a few glasses of water. While not as pervasive as code, food and water seem to be slightly more important. And yet we do not promote plumbers and farming. We don't even promote basic home maintenance and cooking. But coding is a requirement to "keep up with the world" and should be taught in schools, and cooking is not.

In this way coding is much like a language. You may learn a coding language just as a second language. But the value is not in the language that you speak, but in the fact that you are able to speak and therefore communicate information. The value of coding is not in the language you learn to code in, but in the fundamental problem-solving skills that coding requires.

But those skills exist in literally everything. The algorithm for turning a piece of work on a lathe to create a vase is no different from the algorithm for picking the lowest price from a database. (Seriously, they are both optimization search problems with one WHILE and an IF statement.) The question is if the individual is able to plan out that algorithm and identify what is needed to make it work. This is why coding is a tool. It is a method of building, testing and demonstrating problem-solving, in a medium that is, frankly, cheaper than wood or metal shop.

Coding is also hyped because it seems like the future We love to imagine and theorize about the future, and more code is definitely in that future. The trouble is you don't get robots and flying cars from programmers alone. You need machinists and mechanical engineers, people who don't always connect with code. But the simple fact that Code is a cheap means of teaching problem solving gives it the advantage over wood or metal shop in the classroom. But that limits the number of people that can engage in learning problem-solving to create things, and creates a glut of STEM brains creating code for the next Facebook instead of designs for the next spaceship. This has lead to sayings such as "We wanted flying cars, instead we got 140 characters" by Peter Theil, billionaire lawyer cofounder of Paypal, Palantir, and Founder's Fund, major investor in SpaceX, and advocate of skipping college in favor of just creating something.

If you like coding, great. It is certainly useful and integral to our world. But I also say, if you like business, great. It is certainly useful and integral to our world. I also say if you like fixing cars, great. It is certainly useful and integral to our world.

The thing that makes a person valuable is not the specific tools that they use in their career. Whether it be code, CAD, or a wrench. It is how well they understand and utilize those tools, and what they can build with them inside of the real world. But to say that code should somehow get the emphasis, is simply wrong. Every twitter programmer has to come home to a house. And someone is going to have to design and build the plumbing on Mars. Programming is simply a trade, just like any other trade.

+++++++++++++++++++++++++++

At first I thought I would be able to collect my thoughts on STEM education all in one post. After several hours of work I realized that that would not be possible. So there will be a few more most like this to follow in the next few weeks. I apologize ahead of time.

3D Printing and the Future of Manufacturing

It has been a long time since I posted here. Let's see how it goes.

I recently have the pleasure of giving a Talk at the TedxBoise event several weeks ago. The name of my talk "What to do with a giant 3D Printing Farm." I will post the video when it is dropped but the talk was primarily about how I came to be so heavily involved with 3D printing over the last 1-2 years, after having dismissed it for most of my career.

I'm going to take this post to summarize and clarify some aspects that of 3D printing that I wasn't able to speak about. Hopefully this post will be a complementary resource to the video when it drops.

So what do I think of 3D printing? It is a ridiculous personal technology. Owning a personal 3D printer is like owning a personal Bandsaw. Everyone can use them, very few can use them well because it actually does require skill. Personal 3D printing is a failed idea, because physics and the non-scientific way that it was created makes it impossible to have the reliability that a home appliance needs. To show this let me ask you a question. When was the last time you replaced a non-disposable component of one of your appliances. The faucet on your sink? The rotor in your Dishwasher? The most involved with the appliances we use is in replacing the capsule in a Keurig or the cartridge in our Paper Printer. But a 3D printer will have a nozzle clog, because physics, and when that happens you have to identify the problem and then have the tools and wherewithall to replace the nozzle. That disqualifies personal 3D printing as a personal technology. And no amount of software will ever prevent a nozzle from clogging. I will leave that there. I could go on, but my beef with personal printing is long and comprehensive.

The requirement of 3D printing of skill to operate it means that utilization of 3D Printing by regular people will be through services. Shapeways, 3D Hubs, etc. But through several projects that I was involved in I found that one service of 3D printing was missing. Production.

Since 3D Printing is a digital process with nearly no geometry constraints it can be used to make almost any piece of hardware. Now most will counter that statement with 3 basic arguments. 3D printing is Slow, Crappy, and Expensive. These are actually the core of my Tedx Talk so, to avoid spoilers at this point, suffice it to say that each of these are wrong and have been proven wrong by one of the largest 3D printing factories on the planet that I have the pleasure to be a founder of.

3D printing is a viable production method. But it is only viable up to a certain point. On average at Slant 3D production with 3D printing is more affordable than injection molding up to about 20,000 parts (give or take). We plan to make it cheaper up to 100,000 or more. At that point the capital expenditure that makes molding prohibitive basically drops away in the scale. The cost of the part is in the cost of the material and the energy that goes to making it. Oddly this is true of 3D printing. So once there are facilities large enough to produce millions of pieces with printing, there will be no reason to use molding.

The caveat there is that molding is reliable and very fast at very high volumes. I don't think that printing will ever replace it at that scale. But it will start absorbing the low volume production of startups and inventors. The digital nature of it lets inventors design and iterate on 10 different version of a product for very little cost. The same way you would with software. This hasn't quite become mainstream yet because there are really only a 2-3 companies in the world with the 3D printing capacity to take a design from prototype to production entirely with printing. Slant 3D is one of them.

But as 3D Printing manufacturing becomes more mainstream I believe that large scale production will decrease and be supplanted by niche products. Note niche here does not mean ultra-customized. Customization is also a black hole. But niche markets of several thousand clients. As more people come into this space the need for "homerun" products will decrease and the production methods to create them will not be as necessary. Only truly ubiquitous products would be created with mass production. Phones, lights, clothes. But toys, tools, accessories can all parsed out to individual groups and produced with giant 3D printing farms.

3D Printing is the future of "manufacturing" not the future of personal manufacturing, Amazon same day delivery invalidates that. But Amazon will always need new products to delivery.

The work we're doing at Slant 3D for the entrepreneur in the dorm room with a CAD program not a compiler.
_______________________________________

Its good write again. I think I will do this more often. I have had several writing projects on hold, I think I will be able to get back to them.